By the way, it isn't that it gets the focusing wrong but if you try again it gets it right. It gets "stuck" and once it has failed repeated attempts fail. Mostly. But occasionally it flips back again. It is not consistent.
Hmmmm, low contrast subject, high contrast background - but you would think it would focus on the boundary line between the two.
Focus on something nearer first, then swing onto subject?
Repeat with FZ200 and 70D?
Mike
Since we are talking about birds in flight (sort of) it makes sense to use phase detect focusing with the 70D. I tried tests somewhat like I did with the FZ330 and FZ200, but couldn't work out a way of letting you see what I saw through the viewfinder. In an attempt to do that I rigged up the 70D's HDMI output to an external screen. Unfortunately this doesn't show what is on the OVF. It only shows what is on the LCD when using live view.
However, it turns out that it looks and acts the same as the OVF if you use the 19 point phase detect approach in live view on the LCD. (It isn't actually the same. For example, it seems to need more light than the OVF version. In live view it must be doing phase detect off the sensor - most of its ("dual pixel") pixels are phase detect capable, and as the mirror is permanently up in live view it can't be using off-sensor phase detect. So it is emulating what happens through the OVF, and its 19 points in live view act very similarly to the 19 points in the OVF. But not quite so sensitive it seems.)
So I photographed what I was seeing on the external screen. (Had I thought about it for a second or two I would have realised that I could have simply disconnected the HDMI output and the LCD would have been live again, and I could have photographed that like I did for the FZ330 and FZ200.)
Here is the setup. Despite appearances in this photo, it wasn't dark yet, and focusing worked fine with the OVF, but not with live view. I had to add some light to the scene. There were three targets at different distances in front of the books in the background - two piglets and a bottle of eyedrops on a steadying stick.
FWIW, and irrelevant as it is, here is the setup on the camera side of things.
The bottle of eyedrops was closest to the camera. If the bottle was anywhere (ish, see below) in the focusing area, which is marked out by the whiteish lines, that is where the camera would focus. Or on the stick, even though the stick was just a little further away. Perhaps because it was brighter? Don't know.
Next, I've moved the camera but the stick is still in the focusing area. It doesn't get picked up though, and the next nearest target is focused on.
Now only the more distant piglet is in the focusing area. That is where the focus falls.
Now the stick is back in the focusing area, and away from the edge. The focus falls on the stick.
The stick is now on the left edge. Like on the right edge it doesn't get picked up. The background is picked up. It looks a bit brighter at that end of the books, but I suspect it focused there because the angle of the bookcase means that it is nearer on the right.
Now the stick is back on the left, a bit further in this time, but it's still not getting picked up. The back of the chair gets picked up instead.
The stick is further in now and it gets picked up.
So, phase detect focusing knows about distance and picks the nearest target that it can make out. This would explain why the 70D does so much better than the FZ330 with subjects that are in front of a high contrast background. It isn't perfect, if the nearest thing is only visible at the edges, but other than that it seems pretty clear what is going on, and is definitely a better deal for birds in flight. IMO.
--
Nick
The safest and most inclusive global community of photography enthusiasts. The best place for inspiration, connection, and sharing!
www.flickr.com
GardenersAssistant Photography Videos -
https://www.youtube.com/channel/UCmBgEwRDfiQMYTPORSzDxvw